Can large language models enhance assessment of Criterion A for PTSD from self-report?

Key Points

• Enhanced reporting of PTSD Criterion A events via self-report improved when using LLMs, leading to increased information.
• Participants reported more events meeting Criterion A while using LLMs, though accuracy still lagged compared to clinicians.
• Observational analysis on the integration of LLMs in treatment and online research for PTSD assessments was conducted.
• These findings highlight the necessity for further exploration of LLM applications in psychological assessments.